Parallel POCS Interpolation for Seismic Deblending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing POCS interpolation algorithms for seismic data are computationally expensive and time-consuming, limiting their application to datasets that could benefit from improved accuracy and efficiency.
Innovation Solution
A modified POCS interpolation algorithm that parallelizes thresholding across all frequency bands, reducing the number of iterations by a factor of 10 or more, while maintaining accuracy by utilizing maximum threshold values generated from every individual frequency band.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional POCS interpolation algorithm is used, then interpolation accuracy is achieved, but processing time and computational cost are excessively high
Solution Approach 1:
The patent segments the frequency spectrum into multiple frequency bands and processes each band independently through parallel computation. This divides the large-scale computational problem into smaller, manageable sub-problems that can be solved simultaneously, reducing overall processing time while maintaining interpolation accuracy across the full frequency range
Solution Approach 2:
The patent introduces parallelization across the frequency band dimension, transforming a sequential single-band processing approach into a parallel multi-band approach. By adding this computational dimension, the algorithm processes multiple frequency components simultaneously rather than sequentially, dramatically reducing iteration time
2Measurement precision
If traditional POCS interpolation algorithm is used, then interpolation accuracy is achieved, but computational resources required are excessive
Solution Approach 1:
By segmenting the computational workload into independent frequency band processing tasks, the patent enables more efficient utilization of computational resources. Each segment operates independently with reduced memory and processing requirements compared to handling the full spectrum sequentially, lowering overall resource consumption while achieving the same interpolation accuracy
Solution Approach 2:
The patent merges multiple parallel frequency band processing operations into a unified interpolation framework. By combining the results from parallel processing of multiple frequency bands, the algorithm achieves complete spectrum reconstruction with reduced total computational resource requirements compared to sequential processing
3Productivity
If simultaneous source acquisition with blending is used, then data acquisition efficiency is improved, but data separation complexity increases
Solution Approach 1:
The patent applies local quality by processing each frequency band with tailored thresholding parameters and coherence criteria specific to that band's characteristics. This localized processing approach simplifies the overall deblending problem by addressing frequency-specific signal properties rather than applying a uniform complex algorithm across the entire spectrum
Data Source
AI summary
Seismic data may provide valuable information with regard to the description such as the location and/or change of hydrocarbon deposits within a subsurface region of the Earth. The present disclosure generally discusses techniques that may be used by a computing system to interpolate or deblend data utilizing a projection on convex sets (POCS) interpolation algorithm. The utilized POCS interpolation algorithm operates in parallel for frequency of a set of frequencies of a seismic frequency spectrum.